CVE-2022-29212: TensorFlow quantization assert

CISO Take

A crafted TFLite quantized model can crash the TFLite interpreter via an assertion failure in the quantization scaling logic. If your deployment loads externally-sourced or user-provided TFLite models, this is a denial-of-service vector. Patch to TF 2.9.0, 2.8.1, 2.7.2, or 2.6.4 immediately and audit model ingestion pipelines for untrusted inputs.

What is the risk?

Medium risk overall, but context-dependent. CVSS 5.5 reflects local attack vector and availability-only impact. However, in edge AI or mobile inference deployments that accept third-party TFLite models, this becomes a reliable DoS: low complexity, no user interaction after model load. The real exposure is in model marketplaces, federated learning pipelines, or any system where the model file originates outside the organization's control.

What systems are affected?

Package	Ecosystem	Vulnerable Range	Patched
TensorFlow	pip	—	No patch
195.8K OpenSSF 7.1 3.7K dependents Pushed 2d ago 4% patched ~1372d to patch Full package profile →

Do you use TensorFlow? You're affected.

How severe is it?

CVSS 3.1

5.5 / 10

EPSS

0.3%

chance of exploitation in 30 days

Higher than 23% of all CVEs

Source: EPSS v3 — FIRST.org

Exploitation Status

Exploit Available

Exploitation: MEDIUM

Sophistication

Moderate

Exploitation Confidence

medium

○ CISA SSVC: Public PoC

○ Public PoC indexed (trickest/cve)

Composite signal derived from CISA KEV, VulnCheck KEV, CISA SSVC, EPSS, Metasploit, Exploit-DB, trickest/cve, Nuclei templates, and inthewild.io exploitation reports.

What is the attack surface?

AV Local

AC Low

PR Low

UI None

S Unchanged

C None

I None

A High

What should I do?

5 steps

Patch: Upgrade TensorFlow to 2.9.0, 2.8.1, 2.7.2, or 2.6.4.
Model validation: Implement cryptographic signing and integrity verification for all TFLite models before loading — reject models from untrusted sources.
Sandbox: Run TFLite inference in isolated processes so a crash does not take down the parent service.
Detection: Monitor for abnormal process termination or assertion failures in TFLite inference services (SIGABRT signals).
Inventory: Audit which services load TFLite models and from what sources, prioritizing externally-sourced model pipelines.

What does CISA's SSVC say?

Decision Track*

Exploitation poc

Automatable No

Technical Impact partial

Source: CISA Vulnrichment (SSVC v2.0). Decision based on the CISA Coordinator decision tree.

How is it classified?

DoS Supply Chain Framework Inference Model AML.T0010.001 - AI Software AML.T0011.000 - Unsafe AI Artifacts AML.T0029 - Denial of AI Service

Which compliance frameworks are affected?

This CVE is relevant to:

EU AI Act

Article 9 - Risk Management System

ISO 42001

8.4 - AI System Lifecycle — Operation and Monitoring

NIST AI RMF

GOVERN-1.4 - AI Supply Chain Risk Management MANAGE-2.2 - Risk Treatment and Response

Frequently Asked Questions

What is CVE-2022-29212?

A crafted TFLite quantized model can crash the TFLite interpreter via an assertion failure in the quantization scaling logic. If your deployment loads externally-sourced or user-provided TFLite models, this is a denial-of-service vector. Patch to TF 2.9.0, 2.8.1, 2.7.2, or 2.6.4 immediately and audit model ingestion pipelines for untrusted inputs.

Is CVE-2022-29212 actively exploited?

Proof-of-concept exploit code is publicly available for CVE-2022-29212, increasing the risk of exploitation.

How to fix CVE-2022-29212?

1. Patch: Upgrade TensorFlow to 2.9.0, 2.8.1, 2.7.2, or 2.6.4. 2. Model validation: Implement cryptographic signing and integrity verification for all TFLite models before loading — reject models from untrusted sources. 3. Sandbox: Run TFLite inference in isolated processes so a crash does not take down the parent service. 4. Detection: Monitor for abnormal process termination or assertion failures in TFLite inference services (SIGABRT signals). 5. Inventory: Audit which services load TFLite models and from what sources, prioritizing externally-sourced model pipelines.

What systems are affected by CVE-2022-29212?

This vulnerability affects the following AI/ML architecture patterns: model serving, edge/mobile inference, training pipelines, model distribution pipelines.

What is the CVSS score for CVE-2022-29212?

CVE-2022-29212 has a CVSS v3.1 base score of 5.5 (MEDIUM). The EPSS exploitation probability is 0.32%.

What is the AI security impact?

Affected AI Architectures

model servingedge/mobile inferencetraining pipelinesmodel distribution pipelines

MITRE ATLAS Techniques

AML.T0010.001 AI Software

AML.T0011.000 Unsafe AI Artifacts

AML.T0029 Denial of AI Service

Compliance Controls Affected

EU AI Act: Article 9

ISO 42001: 8.4

NIST AI RMF: GOVERN-1.4, MANAGE-2.2

What are the technical details?

Original Advisory

TensorFlow is an open source platform for machine learning. Prior to versions 2.9.0, 2.8.1, 2.7.2, and 2.6.4, certain TFLite models that were created using TFLite model converter would crash when loaded in the TFLite interpreter. The culprit is that during quantization the scale of values could be greater than 1 but code was always assuming sub-unit scaling. Thus, since code was calling `QuantizeMultiplierSmallerThanOneExp`, the `TFLITE_CHECK_LT` assertion would trigger and abort the process. Versions 2.9.0, 2.8.1, 2.7.2, and 2.6.4 contain a patch for this issue.

Exploitation Scenario

An adversary targeting an organization's edge AI inference service (e.g., a computer vision pipeline on IoT devices or mobile apps) crafts a TFLite model with quantization parameters that include a scale value >= 1.0. They publish this model to a public model repository or submit it via a model upload feature. When the vulnerable TFLite interpreter loads the model, the TFLITE_CHECK_LT assertion in QuantizeMultiplierSmallerThanOneExp triggers SIGABRT, crashing the inference process. In a fleet deployment scenario, all devices pulling the same model update simultaneously become unavailable — a scalable, low-effort denial-of-service against production AI infrastructure.

Weaknesses (CWE)

CWE-20 Improper Input Validation

CWE-20 — Improper Input Validation: The product receives input or data, but it does not validate or incorrectly validates that the input has the properties that are required to process the data safely and correctly.

[Architecture and Design] Consider using language-theoretic security (LangSec) techniques that characterize inputs using a formal language and build "recognizers" for that language. This effectively requires parsing to be a distinct layer that effectively enforces a boundary between raw input and internal data representations, instead of allowing parser code to be scattered throughout the program, where it could be subject to errors or inconsistencies that create weaknesses. [REF-1109] [REF-1110] [REF-1111]
[Architecture and Design] Use an input validation framework such as Struts or the OWASP ESAPI Validation API. Note that using a framework does not automatically address all input validation problems; be mindful of weaknesses that could arise from misusing the framework itself (CWE-1173).

Source: MITRE CWE corpus.